Speaker
Description
Phylogenetics provides feedback for public health programs’ effectiveness, including for pre-exposure prophylaxis (PrEP), a key pillar to end the HIV epidemic. In British Columbia (BC), PrEP has been provided free of charge since 2018, but eligibility criteria, access barriers, and adherence diminish its potential. We estimated HIV cases in BC averted through PrEP, which we hypothesized were heterogeneously distributed across phylogenetic clusters. 42,043 HIV partial pol sequences from 10,740 individuals were aligned to HXB2 reference using minimap2 and drug resistance mutations were masked. Bootstrap phylogenies were inferred in FastTree2 to identify phylogenetic clusters with >4 members with pairwise patristic distance <0.02 substitutions/site in >90% of bootstraps. We calculated clusters’ growth rates and effective reproduction numbers (Re) using new diagnoses over time, then averaged over periods preceding (Jan 2016 – Dec 2017) and following (Jan 2018 – Feb 2020) widespread PrEP. Counterfactual models fit to growth rates preceding PrEP were used to quantify cases averted. Of 84 actively growing clusters (new cases since 2018), 44% of 52 clusters predominantly comprised of men who have sex with men (MSM) had significantly lower growth rates following PrEP compared to 17% of 30 clusters comprised of people who inject drugs (PWID); corroboratively, Re was reduced in 45% and 31% of gbMSM and PWID clusters. PrEP averted 57-125 HIV cases between January 2018 and February 2020, which were primarily concentrated in four active clusters of young (median age first ART:27-35) gbMSM (79-92%), and included previous PrEP users who were later diagnosed, indicating PrEP use among cluster contacts. Rapid growth of clusters of PWID and older median age indicated missed opportunities for PrEP. Cluster-level modeling of interventions’ effectiveness revealed PrEP differentially reduced clusters’ growth, but cumulatively may have averted 2.2-4.8 new HIV cases per month. Combining phylogenetics and counterfactual models is a pragmatic approach to inform public health policy.
